computably

Or, you know, automatically, with plain old software. Which works fine and fails in less ridiculous ways.

> I suspect that for most people, much of it is a feeling of accomplishment from performing a difficult skill that creates something of value for other people. It's the joy of being known as a good cook, and the satisfaction of making something for people that they might otherwise have trouble getting. How much fun would cooking be if every street corner had a Michelin-starred restaurant giving food away for free? How much would you cook if none of your friends really cared about the results because they can get better food elsewhere, cheaper?

I think it's paradoxical to suggest Michelin-starred restaurants could be everywhere, given the whole point is the travel-worthiness i.e. a scarcity that justified effort.

That aside, to take the strongest form of the argument, I think even if every human being could get their ideal meal, no matter what it is, at any time and for free - and even if the cook in question is only doing it for social reasons - cooking would still be valued. It's the effort put in by a person which is scarce and precious. Even if people were immortal, that time in the present would still be valuable in itself.

> How is kindness demonstrated in a world where no one wants for anything?

Post-scarcity of material things won't sedate humans. Kindness is a thing wanted in and of itself, but more importantly, internal struggles and desires can't all be resolved through material gain.

> How can someone be more interesting than synthetically generated content created specifically attuned to our personal interests?

The answer is apparently, pretty easily? By and large it seems the consensus is social media and digital content are exploiting a niche of providing a low-effort, highly-addictive pastime.

If the service provider goes under, the govt wasted their money.

Reviewers are also usually biased towards reviewing products that are good and/or interesting, and while that's also true of consumers, you'd expect reviewers to be more informed.

What you're describing is a language that can effectively take a high-level design doc (limited to a formal language representing a pseudo-English-subset) and compile to almost arbitrarily complex, formally verifiable systems. I say systems because you're also implying these could essentially orchestrate distributed systems.

Which, if that were really possible, would be pretty good evidence of AGI. Or, at least, the end of 6-7 figure software engineer salaries. Conversely, I think it's a pipe dream. Writing a compiler that actually works well across such a broad range of hugely complex behaviors would itself be absurdly complex.

> If 90% of new games get tagged as AI no matter how non-slop they are

You presume in favor of the existence of non-slop.

> There as good games that use AI

[citation needed]

There theoretically could be - and I don't have an issue with an otherwise competent dev/artist making use of AI tools - but I know no examples of good games that also deserve a "made with AI"/"uses AI" badge.

YT's business model is monetizing creators' output. Producing a video that gets a million views and generates a bunch of ad revenue is the fee.

> I used to think that compilers should use more verification technology, but people start screaming about compile times.

"Just" make verification faster. ;)

I assume they're referring to some amplification of that latency that makes it easily observable. e.g. multiple round trips on a user interaction

That's a false dichotomy. An AI assistant doesn't have to be unrestricted and monolithic.

From the perspective of the humans who first harnessed fire, it could have been a humanity-ending threat for all they knew. Certainly for the atomic bomb, they briefly entertained the possibility it would literally ignite the entire atmosphere.

And under a long enough time horizon, you could just as well argue that pretty much every scientific advancement has contributed to the development of modern AI. Does it count as AI doom if a century from now humans just start a plain old World War N with barely-intelligent killer robots?

That's loaded with ambiguous qualifiers. Anybody could say that about pretty much any revolutionary technology. Unless you have a convincing argument for how there's literally only one way for it to go right, it's an unfalsifiable claim.

There have always been Adobe alternatives and many are more than 50% functional.

I'm pretty sure the "proprietary" part is the gatekeeping.

Depends on your definition of doom.

If doom is ending up with grey goo / paperclip maximizers, or SkyNet, then I don't think doing mathematics is evidence of that direction. Partly because LLMs are quite apparently dumb in many ways, and for math specifically, they need a formal verifier (Lean) which "gamifies" math.

If you mean bioweapons or cyberwarfare, there's nonzero risk, but not orders of magnitude worse than other global risks. Climate change, nuclear weapons, monoculture food, etc.

I'm far more concerned about overall trends in AI development and usage. It's accelerating wealth inequality, social isolation, attention capture, surveillance states. If we end up in the Matrix except the admins are humans and the simulation is hyperoptimized TikTok, is that AI-driven doom, or is it just an inevitable outcome of modern tech?

Huh? The main benchmark is compressed size + program size. The uncompressed size of enwik9 is defined as 1GB, so no "L"LM qualifies.

I doubt that agents can make a product as polished as Slack. The status quo would indicate that's not the case as the prediction that SaaS would die has not panned out at all.

Not sure if you're referring to the actual definition. It's basically "no two participants from different pairs in the matching would prefer each other to their current matches." It seems an elegant definition to me.

This obviously doesn't factor in most of the complexity of romantic relationships: not knowing your own preferences, evolving preferences, incomplete/imperfect information in general, etc. Plus it assumes a global 1:1 matching across two categories, and people can of course be LGBT or choose to be single.

I left another comment with the minimal example, but intuitively:

- proposers descend from their 1st choice, while recipients ascend according to their offers; the outcome can't be recipient-biased, because recipients ascend only when proposers are forced to descend

- a matching is stable if each pair contains at least one party that cannot find a strictly better match

- in particular there are multiple distinct matchings, and if proposer-biased, recipients have no recourse to break the proposer-favored pairs

(A,X),(B,Y) as the proposals is fine for the point being made. The issue is if we reverse which group is the proposer, we still end up with the same matching, because (A,X) strictly prefer each other.

The minimal example with 4 participants is if each person P has unique preferences and P's 1st choice has P as 2nd choice. So in this case just flip X:

A:XY B:YX X:BA Y:AB

If A,B propose you get AX,BY. If X,Y propose you get BX,AY. Both are stable because the proposers are getting their first choice.

> There is no reliable way for a dating app to only serve users non-committal matches for example

While that certainly sounds difficult, it sounds feasible for e.g. Facebook.

> The thought experiment I usually pose is to ask what you think a free, open source, and community supported dating app would change. The answer is "not much".

That sounds more like a framing problem. If the ultimate goal is promoting long-lasting new romantic relationships, and it's developed without any profit motive, I'd imagine a general social network that primarily organizes in-person social gatherings.

Not sure if you're trying to be humorous but RSI is when a system improves itself. It's explicit in two out of the three words...

Sure, I acknowledge an open ecosystem is at least not something Apple advertises nor what any reasonable consumer believes Apple to be advertising.

My point is that "users installing anything they like" is, or ought to be, a legal right. It's an intrinsic part of ownership of a general-purpose personal computing device.

> It's very shocking to see this turn of the editorial approach for a company that did change the full industry.

Is it? Regardless of how revolutionary a company was, they don't get a free pass on criticism.

I had a positive sentiment towards Tesla when the mission was basically "make EVs sexy" and "promote clean electricity." Although they didn't have a perfect record (QC, proprietary charging, RtR), at least they were disrupting a stagnant industry, with a noble enough goal.

Now when I think of Tesla, I think of their deployment of "Full Self Driving (Beta) [keep your hands on the wheel and eyes on the road]", "Robotaxi (with human driver)," "Humanoid robot (why?)," and of course the glorious leader who bought Twitter so he could "protect free speech" by banning people he disagrees with.

HCI started as a field before the idea of "personal" computers was realized. Now most humans on earth have a computer in their pocket, often with constant access to a global network and infinite addictive feeds of content to consume. I wouldn't call it scope creep, it seems like a natural evolution.

None of that explains why the owner of an Apple device can't install alternative app stores, nor the bundling of their own app store.

If it's about the latency / flow disruption, spending a few hours once could easily be worth it if the result is actually good enough to skip googling/retries.

Even ignoring speed per GP, the Hutter Prize's metric includes the size of the decompressor. LLMs would be disqualified for being larger than 1GB.

> say "better than 95% of humans on 95% of intellectual tasks" but if we used that definition we already have AGI and almost no one thinks we have achieved AGI

What matters isn't 95% of humans, it's 95% of actual professionals. Benchmarking an AI accountant against people with zero accounting experience is worse than worthless.